As for that ridiculous report about the earth being warmer than for thr last 1800 years- it is based on highly discredited techniques and is not at all credible.
"Greenhouse scientists Michael Mann and Philip Jones (GRL v.30, no.15, 1820, 2003) have launched `Hockey Stick Mk.II'. It will be remembered that Hockey Stick Mk.I was published several years ago and became instantly adopted as policy by the IPCC. In a nutshell, the `Hockey Stick' theory presented a 1,000-year `reconstruction' of northern hemisphere temperature and in so doing denied the existence of the Medieval Warm Period and the Little Ice Age as hemispheric or global events during the previous millennium. This denial then allowed them to incorrectly assert that the 20th century was warmer than any previous century and that `1998 was the warmest year of the millennium'. This was in spite of a large body of peer reviewed literature confirming the existence of these two climatic events during the last 1,000 years.
Hockey Stick Mk.II now extends the time period back 1,800 years, but using much the same technique, this time taking a mere 23 selected proxy sites (tree rings mostly) to represent the whole world, processing them statistically, and then grafting the result onto the CRU version of global temperature during the last century (that CRU series is itself disputed due to urban contamination of weather station temperature records).
Is this version of climate history credible? Just because it has appeared in peer-reviewed literature is no sufficient reason to accept it. Papers which affirm the existence of the MWP and LIA are peer reviewed too - and much more numerous. So what might be wrong with Hockey Stick Mk.II? Basically, everything that was wrong with Mk.I, namely it denies a mountain of contrary evidence.
Three central problems arise with Hockey Stick Mk.II.
Firstly, grafting a data stream from one type of marker (tree rings and ice cores) onto a completely unrelated data stream (weather station records, most of them urban) is simply bad practice statistically. It's the proverbial apples and oranges comparison, making the result quite meaningless.
Secondly, there was no attempt to deal with the huge volume of peer reviewed literature attesting to the existence of the two major climatic events which Mann & Jones deny - the Medieval Warm Period and the Little Ice Age. One recent paper by Soon & Baliunas (Clim. Res. 23, 89-110, 2003) also used proxy data and concluded the two events really did happen. But Mann & Jones' response was this comment, delivered as if it were a statement of fact instead of what it really was - merely a statement of their opinion.
"A flawed recent study [Soon and Baliunas, 2003] compels us to stress two points..." etc. etc.
A statement opened and phrased in that way is a clumsy attempt to pre-judge the issue and can only be viewed as a political statement, not a scientific one. Such political tactics do not belong in scientific literature and should have been queried by the reviewers. Opinion is not science, especially in a controversial matter like this.
Finally, tree rings are an unsatisfactory proxy for determining regional or global temperature. They only cover the growing season, not the winters. They only cover the daytimes, not the night. They only cover a fraction of land areas, with no ocean coverage at all (amounting at best to perhaps 15% of the planet's surface). The rings themselves are conditioned by a variety of environmental factors of which temperature is only one.
With such limitations, it is absurd to even imagine that something as grandiose as `global mean temperature' could ever be determined from them, let alone `reconstruct' past temperatures as Mann & Jones have done to a claimed accuracy of tenths of a degree!
As to the 23 proxy sites, that is much too small a sample to conclude anything, particularly as they were selected against other proxies which may have told a different story. If one selectively chose 23 weather stations out of the thousands available, one could even prove we were headed for an ice age, or a global heat wave - selectivity and intensive statistical processing allows any outcome to be produced.
The final nail in the coffin of Hockey Stick Mk.II is contained in their own graphs. In their reconstruction of recent decades (shown below), the authors overlapped the weather station record (red line) with the reconstructed proxy record (blue). And the two make a very poor match. The CRU weather station plot shows much wider temperature variations than the proxy plot. If the weather stations are assumed to be more accurate than the proxies, then the 1,800-year reconstruction presented by Mann & Jones is largely an artifact of statistics and data selectivity. As one commentator remarked -
"It looks like the authors are just averaging together a bunch of noise.""